MEGA-Bench Task - GUI Act Web Single: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-8B | 13.1 |
| 2 | Gemini 2.5 Pro (Preview 03-25) | 9.8 |
| 3 | Qwen 2 VL 72B | 8.6 |
| 4 | Llama 4 Scout Base | 7.8 |
| 5 | Gemma 3 27B (IT) | 7.5 |
| 6 | Claude 3.5 Sonnet (20240620) | 7.1 |
| 7 | Qwen 2 VL 2B | 6.8 |
| 8 | Gemma 3 12B (IT) | 4.8 |
| 9 | Gemini 1.5 Pro (002) | 3.9 |
| 10 | Pixtral-12B | 3.9 |
| 11 | Gemini 1.5 Flash (002) | 3.9 |
| 12 | Qwen 2 VL 7B | 3.3 |
| 13 | Claude 3.5 Sonnet (20241022) | 3 |
| 14 | GPT-4o Mini | 3 |
| 15 | GPT-4o | 1.6 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-gui-act-web-single · How It Works · Data refreshed daily, snapshot 2026-09-05.